Only Truth Matters

1. Intro

ROS is too heavy for new project with little dependence on open-source project.

Plan: Understand the design and philosophy of ROS1 and ROS2, design my own robotics stack

2. ROS1

2.1. Philosophy and Design Principles

Goal: to support code reuse in robotics

• as thin as possible? Really?
• ROS-agnostic
• Scaling
• Test, Language

How: Node for executables to be run individually

2.2. Architecture

Ref

1. Intro

In the past few month, I use ChatGPT intensively, the CEO of OpenAI is Sam Altman, he wrote a blog called 「How to be successful」, today I wanna look at it and see what I can learn from

2. Content

Ref

For Altman, the definition of success is

• Make huge amount of money
• Create something important

Tips

1. Compound yourself
2. Rate of learning should be high
3. Self-belief
4. Trust yourself more
5. Balance with self-awareness
6. Think independently
7. Get original idea
8. Find easy, fast ways to test these ideas
9. Get good at ‘sales’
10. convince other people of what you believe
11. Show up in person whenever its important
12. Take risk
13. Its impossible to be right all the time
14. Human nature, prioritize short-term gain
15. Focus
16. Much more important to work on the right thing
17. Work hard
18. do it at the begining of your career
19. Be bold
20. people want to be part of something exciting and feel the work matters
21. Follow your curiosity
22. Be willful
23. people have the power to change the world
24. Be hard to compete with
25. do what others can’t
26. personal relationships
27. strong personal brand
28. good at intersection of multiple different fields
29. Build a network
30. the size of network become the limiter for what you can accomplish
31. Get rich by owning things
32. don’t sell time
33. Be internally driven

1. Goal

因为头发长得比较快而且是短发类型，需要经常修剪。

去理发店耗费大量时间精力，充满不确定因素。

自己剪头发，熟能生巧，逐渐掌握适合自己的发型，且随时修改，方便快捷

2. Tool

自己理发需要一些工具，不是必须但是有的话能大大提升效率

• 普通剪刀：平剪，用于剪掉100%的头发
• 牙剪（重要）：有不同去量型号，比如15%，30%等等，就是每次剪下去会有15%或者30%的头发被剪掉，我在给自己的理发过程中主要使用这一工具，因为每次去量较少，所以结果不具有破坏性
• 多面镜子（重要）：有专门的理发镜子，三折，通过这种方式可以看到自己侧边和后边的发型，
• 电动推子：配合不同长度的头，可以达到精准控制长度的效果
• 吸尘器：自己理发会有很多头发掉落在身上以及地上，理完发及时清洁可以避免碎发跟随脱衣或衣物扩散至其他地方

3. Principle

基本原则是：

• 最重要：在不确定效果时，少量修改，如果看到效果不满意就换种方式。基于Feedback的理发方式，早起频率可以在1天一次，熟练之后可以一周一次
• 目的：达到自己想要的发型，只要自己看着顺眼即可，所以需要先观察，观察自己当前的发型有哪些不满意的地方，比如说xx地方突出来了，xxx地方太平整，没有层次
• 胆大心细：一开始的时候总会有剪错的时候，发现发型的问题之后尝试解决方案，用牙剪可以避免重大失误。效果不好就及时停止，并记住错误的方式，等头发恢复一点再尝试其他方式即可

1. Vision For a Better World

How to do that (in general)? Find how current world behave bad, (not ideal from my view). fix it

Example:

• People do too much work or thing for a living, so they don’t have time for thins they really enjoy
• Imbalance of wealth
• Unfair between people, power uneven distribution
• Environmental issue

How to solve the problem?

1. Robot and Artificial Intelligence should operate the main thing (physical and digital world). For example, the farms, power plant, and other facilities
2. Push the governments distribute enough wealth to everyone

Main Issue

• The selfish nature of human being, make it hard to achieve real equal and fair
• While pushing robot to replace human worker, there would be unemployment issue before the productivity achieve new stage

Practical Step

• Promoting automation
• re-skilling and education
• regulation and safety
• Weath distribution
• Fair Taxation
• Encourage social entrepreneurship
• Fairness
• Transparency
• Access to opportunities
• Environmental Issue
• Sustainable pracitices
• Influence Policy

1. Goal

Understand the general framework of manipulator.

From my current understanding, it would include

• Design (general shape etc)

• Control

• Mechanics

How human manipulate? (manipulation without grasping)

• Sliding and environmental contact
• Grasping

For robot

• Grasping
• Non-Grasping

2. From MIT Course

Ch. 8 - Manipulator Control (mit.edu)

1. Conclusion

1. Interdisciplinary Approach: Many of these seminars showcase the interdisciplinary nature of robotics, highlighting the combination of engineering, computer science, and even biology and physics. Examples include “Snakes & Spiders, Robots & Geometry” by Ross L. Hatton, “Robots and Biosystems” by Kevin Lynch, and “DNA and gammaPNA in programmable nanomaterials…” by Rebecca Taylor.
2. Artificial Intelligence and Machine Learning: A significant number of these seminars deal with artificial intelligence (AI) and machine learning (ML) in the context of robotics. Talks like “Learning to Generalize beyond Training” by Deepak Pathak, “Deep Learning for Robotics” by Pieter Abbeel, and “Machine Learning and Model Predictive Control for Adaptive Robotic Systems” by Byron Boots reflect the current focus of the field.
3. Human-Robot Interaction (HRI): Many seminars discuss aspects of how robots interact with humans, either physically or in terms of communication. For example, “Towards Robust Human-Robot Interaction” by Stefanos Nikolaidis, “Enabling Grounded Language Communication for Human-Robot Teaming” by Thomas Howard, and “Human-Robot Interactive Collaboration & Communication” by Heni Ben Amor, highlight the significance of HRI in modern robotics.
4. Space and Field Robotics: Some talks focus on the use of robotics in space or fieldwork applications. “Mars Robots and Robotics at NASA JPL” by Vandi Verma and “Robotic Grippers for Planetary Applications” by Aaron Parness exemplify space robotics, while “Autonomous and Intelligent Robots in Unstructured Field Environments” by Girish Chowdhary and “The World’s Tiniest Space Program” by Zachary Manchester underscore field robotics.
5. Design and Control: Several seminars delve into the design and control of robotic systems, which can range from the microscale to the macroscale, or even complex, autonomous vehicles. Examples include “Design and control of insect-scale bees and dog-scale quadrupeds” by Avik De, “Design, Modeling and Control of a Robot Bat” by Seth Hutchinson, and “Modeling, Design, and Analysis for Intelligent Vehicles…” by Chung-Wei Lin.
6. Perception and Environment Interaction: Seminars such as “Next-Generation Robot Perception…” by Luca Carlone, “Understanding the Physical World from Images” by David Fouhey, and “Perception-Action Synergy in Uncertain Environments” by Jing Xiao highlight the importance of perception and the ability of robots to understand and interact with their environments.
7. Sustainability and Social Impact: A few seminars also touch upon topics related to sustainability and social impact, like “Robots Should Reduce, Reuse, and Recycle” by Chelsea Finn and “What (else) can you do with a robotics degree?” by Nidhi Kalra.

2. List

Vandi Verma : Mars Robots and Robotics at NASA JPL

Brenna Argall : Mobility and Manipulation Independence with Interface-Aware Robotics Intelligence

Phillip Isola : Structures and Environments for Generalist Agents

Luca Carlone : Next-Generation Robot Perception…

Lerrel Pinto : A Constructivist’s Guide to Robot Learning

David Fouhey : Understanding the Physical World from Images

Jorgen Pedersen: RE2 Robotics: from RI spinout to Acquisition

1. Conclusion

1. Reinforcement Learning and Control Systems: Multiple seminars focus on reinforcement learning (RL) and control systems. Topics such as “Balancing Reinforcement Learning with Model-Based Control”, “Combining Learning and Control in Cyber-Physical Systems”, and “Reinforcement Learning: Leveraging Deep Learning for Control” illustrate the growing importance of RL in the development of intelligent and adaptable robots.
2. Human-Robot Interaction (HRI): HRI is a significant aspect, with seminars like “Towards Human-Friendly Robots”, “Get in Touch: Tactile Perception for Human-Robot Systems”, and “Towards Robust HRI: A Stochastic Optimization Approach”. This suggests a focus on creating robots that can safely and effectively interact with humans in various environments, from homes to workplaces.
3. Robotic Manipulation and Movement: There’s a strong emphasis on enhancing the physical capabilities of robots. Seminars such as “Dynamic Legged LocoManipulation”, “Ballbots”, “Robot Motion Planning”, and “Mechanical Intelligence in Robotic Manipulation” suggest ongoing research to improve robot agility, dexterity, and mobility.
4. Medical Applications: Seminars like “From Tool to Assistant: Towards Developing Adaptive Surgical Robots”, “Tentacle-like Continuum Robots for Minimally Invasive Surgery”, and “Uniting Robots and Ultrasound for Cardiac Repair” highlight the growing role of robotics in healthcare. This trend points to the potential of robotics in augmenting human capabilities in the medical field.
5. Robotics in Industry: Industrial applications of robotics are evident in seminars such as “Robotics and Warehouse Automation at Berkshire Grey” and “University of Michigan’s Work Toward Autonomous Cars”. The trend underscores the transformative potential of robotics in various industries.
6. Bias and Ethics in Robotics: The seminar on “Understanding and Mitigating Bias in Vision Systems” points towards an increasing focus on the ethical implications of AI and robotics, particularly in how systems are designed and how they interact with the world.
7. Autonomous Systems: Autonomous navigation, perception, and control are also key areas, as seen in seminars such as “Robust Control Tools for Validating UAS Flight Controllers”, “Data-to-Decisions for Safe Autonomous Flight”, and “Future of Autonomous Vehicles”. These topics reflect the ongoing research on developing systems capable of operating with minimal human intervention.
8. Career Perspectives: There are also seminars discussing the career trajectories in the field of robotics, such as “Career Options in Robotics: Academia vs Industry”. This trend highlights the growing opportunities and diverse pathways within the field.

2. List

Dynamic Legged LocoManipulation: Balancing Reinforcement Learning with Model-Based Control Towards Human-Friendly Robots: We need more Robots at Home Sort robots for humanity Making Large Dimensional Problems Small Again The Right Stuff: Representing Safety to Get Robots Out in the Real World Combining Learning and Control in Cyber-Physical Systems Get in Touch: Tactile Perception for Human-Robot Systems Robust Control Tools for Validating UAS Flight Controllers Composable Optimization for Robotic Motion Planning and Control Ballbots Insect Scale Multifunctional Micro-Aerial-Robots Powered by Soft Artificial Muscles Field Robotics and Automation at GeorgiaTech Lorraine: An overview Synthesizing & Guaranteeing Robot Behaviors Resilience of Autonomous Systems: A Step Beyond Adaptation Wearable Device Design and Biomechanical Research at Sandia National Labs Robot Motion Planning: Challenges and Opportunities for Increasing Robot Autonomy From Tool to Assistant: Towards Developing Adaptive Surgical Robots for the Operating Room Tentacle-like Continuum Robots for Minimally Invasive Surgery Learning to Walk and Navigate on Legged Robots Understanding and Mitigating Bias in Vision Systems Robotics and Warehouse Automation at Berkshire Grey Shedding Light on 3D Cameras Conflict-Aware Risk-Averse and Safe Reinforcement Learning: A Meta-Cognitive Learning Framework Towards Robust HRI: A Stochastic Optimization Approach Institute for Robotics and Intelligent Machines (IRIM) Industry Day Panel - Career Options in Robotics Career Options in Robotics: Academia vs Industry Robot Learning: Quo Vadis? Reinforcement Learning: Leveraging Deep Learning for Control From Coexistence to Collaboration: Towards Reliable Collaborative Robots Star Wars: The Rise of Robots and Intelligent Machines Cost of Transport, the Correct Metric for Mobile Systems? Navigation and Mapping for Robot Teams in Uncertain Environments Learning from the Field: Physically-based Deep Learning to Advance Robot Vision in Natural Environments Toward Dynamic, Tactical, Remote Robotic Ops: Active Perception and Other Key Technologies Robophysics: Physics Meets Robotics Uniting Robots and Ultrasound for Cardiac Repair Merging Humans and Machines to Assist Legged Locomotion Improving Multi-Fingered Robot Manipulation by Unifying Learning and Planning Considerations on Productivity, Performance and Ergonomics of Human-Centered Robots Deception, ExoNets, SmushWare, and Organic Data: New Frontiers In Neuro-Rehabilitation Safety of Autonomous Systems: Challenges in Coordination, Consistency, and Integration Doing for Our Robots What Nature Did for Us Mobile Data Collection in an Aquatic Environment: Cyber Maritime Cycles for Distributed Autonomy Robotic Manipulation: A Broadening View Human Teacher’s Perception of Teaching Methods for Machine Learning Algorithms Robots for Physical Interaction Factor Graphs for Flexible Inference in Robotics and Vision Robotic and Wearable Sensing Technologies for Movement Rehabilitation After Neurologic Injury Data-to-Decisions for Safe Autonomous Flight Accelerated Optimization in the PDE Framework Mechanical Intelligence in Robotic Manipulation: Towards Human-Level Dexterity in Robotic and Prosthetic Hands Mathematics and Learning for Agile and Dynamic Bipedal Locomotion Robotic Skins That Turn Inanimate Objects Into Multifunctional Robots Creating Robots That See Deep Learning to Learn Geometry and Mechanics of Feet and Fins Soft Miniature Mobile Robots Robots with Privacy Stipulations Physics-Based Manipulation With and Around People Preparing for the Coming Machine Revolution Computational Surgery: Helping Surgeons Avoid Mistakes with Better Robots Vision Based Navigation and Tracking with Small UAVs Developing Sensing and Robotics Technologies for Plant Phenomics The Evolution of Soft Robotics Cyber Human Interaction: A Control Systems/Robotics Perspective on Functional Electrical Stimulation Autonomous, Agile, Vision‐Controlled Drones: From Frame to Event Vision Stochastic Models in Robotics Snakes and Spiders, Robots and Geometry Examining the Slow, Noisy, and Complex Process of Technology Adoption Shared Control of Functional Electrical Stimulation and an Electric Motor in a Hybrid Neuroprosthesis Enhancing Human Capability with Intelligent Machine Teammates Persistent Environmental Monitoring: Robots That Seemingly Do Nothing Most of the Time Behind the Scenes: Decoding Intent from First Person Video Supersizing Self-Supervision: Learning Perception and Action Without Human Supervision Mixed-Integer Convex Formulations for Planning Nonlinear Dynamics in Complex Environments Future of Autonomous Vehicles Closing the Gap Between Machine Learning and Robotics Robust Agility and Safety for Dynamic Aerial Manipulation and Legged Locomotion Weakly Supervised Learning from Images and Video Beyond Geometric Path Planning: Paradigms and Algorithms for Modern Robotics Toward Robots that Understand People and Their Environment Challenges and Opportunities within Maritime Autonomy and the Naval Surface Warfare Center, Panama City Division Robot-assisted Surgery: A Platform for New Technologies and Algorithms University of Michigan’s Work Toward Autonomous Cars The Manipulation Action Grammar: A Key to Intelligent Robots Interactive Sports Analytics: Going Beyond Spreadsheets Multi-agent Collaboration in Shared Workspace Decision Making in Robots and Animals Stochastic Control: From Theory to Parallel Computation and Applications From (Inverse) Optimal Control to Joint (Cooperative) Sequential Manipulation and Motion Planning Towards an Integrated Architecture for Open-World Human-Robot Interactions Robotic and Haptic Manipulation Vision-Based Monitoring of Behavioral Disorders Multirobot Coordination: From High-level Specification to Correct Execution Words, Pictures, and Common Sense Multi‐Robot Systems for Monitoring and Controlling Large Scale Environments Risky Robotics: Developing a Practical Solution for Stochastic Optimal Control Humanoids of the Future The Robotic Scientist: Automating Discovery, from Cognitive Robotics to Computational Biology Efficient Lifelong Machine Learning

1. Conclusion

1. Robot-Human Interaction & Collaboration: This theme is prevalent throughout the seminars, emphasizing the need for more natural and efficient interactions between robots and humans. This covers areas like haptic feedback, interaction through touch, remote presence systems, and even the ethical considerations of autonomous systems.
2. Autonomous Systems & Self-learning: There’s a substantial focus on making robots more autonomous and adaptable through the use of learning algorithms. This includes seminars on autonomous vehicles, field-hardened robotic autonomy, self-supervised networks, and machine learning for robots.
3. Manipulation & Grasping: The issue of creating robotic systems that can effectively interact with their physical environment is clearly important. There are numerous seminars focused on designing grippers, in-hand manipulation, and incorporating tactile sensors for better performance in various tasks.
4. Robotic Design & Fabrication: A number of seminars are dedicated to the development and design of novel robotic systems. This includes topics such as designing bioinspired aerial robots, flexible surgical robots, shape-changing displays and robots, as well as pressure-operated soft robotic snakes.
5. Safety & Robustness: There’s a clear emphasis on making robots more robust and safer, whether it’s through perception-based control, modeling, planning, reachability, or learning in uncertain environments. This theme is especially relevant for autonomous vehicles and robots operating in dynamic or challenging environments.
6. Perception & Sensing: Multiple seminars focus on enabling robots to perceive, understand, and make sense of their environments. This includes topics such as spatial perception, tactile sensing, self-supervision for robotic learning, and incorporating semantics into robot perception.
7. Software & Simulation: The transition from open-source to safety-certified software and the use of simulations for testing and training robots are also recurring themes. This also includes discussions on distributive representations and scalable simulations for Real-to-Sim-to-Real with deformables.
8. Applications: There is a range of seminars focusing on specific applications of robotics, from logistics and surgical assistance to ocean exploration and space.

2. Control-Oriented Learning for Dynamical Systems.

The seminar is about “Control-Oriented Learning for Dynamical Systems.” It focuses on the interplay between learning and control, particularly in the context of robotics.

2.1. One or more of the research problems the speaker’s work address

• Difficulty in modeling robotic systems operating under different conditions, such as icy roads, wind conditions, and contact situations.
• The limitations of naive regression for closed-loop control tasks that happen over long time horizons.
• The need for a learning approach that is more adaptable and efficient for closed-loop control.

2.2. What is the novel approach or idea behind the speaker’s solution, Why it works

• The speaker proposes “control-oriented learning,” which conditions the model to perform better in closed-loop control.
• This approach involves three different lines of work:
• Translating the concept of stabilizability into algebraic conditions used as a regularizer for closed-loop control performance during offline learning.
• Structuring the model appropriately offline to yield a useful control close-up controller based on Linear Quadratic Regulator (LQR).
• Using meta-learning to learn a good extension to a nominal model for adaptive control.
• These methods work because they focus on the downstream control objective rather than just the model fitting objective. They also leverage the inherent structure of the systems to enable control design.

2.3. Questions that maybe asked by an unexperienced robotics engineer and Answers

1. Intro

The apartment charge 2400 for painting

It’s un-reasonable

2. Process

1. Determine Who to Sue
2. Send a Demand Letter (optional but recommended)
3. File Your Case
4. Serve the Defendant
5. Go to Court
6. Collect Your Money

I hope this email finds you well.

In accordance with California Civil Code §1950.5, I am formally requesting the return of my security deposit that I believe has been improperly withheld following my vacating the premises at 833 W El Camino Real, Sunnyvale, CA on June 12, 2023.

Please find attached a letter detailing my concerns regarding an excessive $2400 charge for repainting the apartment, and the lack of an appropriately itemized statement describing the work performed, the time spent, and the reasonable hourly rate charged.

The letter outlines the steps I am prepared to take if we are unable to resolve this dispute, including pursuing a small claims action.

I am open to discussion in order to reach a fair resolution to this matter. I kindly ask you to review the attached letter and provide me with your feedback at your earliest convenience.

1. Intro

Building a SLAM system based mainly on Lidar

3 main different system

• Building point cloud map
• Global localization
• Relative localization

2. Building Map

Graph optimization + loop closure detection

3. Global Localization

Bound and refine bound

4. Relative Localization

NDT (normal distribution transform)